| Main Archive Page > Month Archives > spamassassin-dev archives |
On Mon, 12 Dec 2011, Axb wrote:
> On 2011-12-12 23:45, Kevin A. McGrail wrote:
>
> I've tried to work on some static scores for them, but I've lacked the
> patience to go thru the whole lot. It's huge and good but hard to follow)
Heh.
They basically all work off the same meta list of subrules. _2 is 2+ hits,
_3 is 3+ hits, etc. Then to those some FP-avoidance checks are added.
There are variants for N hits plus a LOTSA_MONEY hit or a FILL_THIS_FORM
hit or both, but in all there are only about twelve variants to score. The
meta list of subrules is generated by a GA rule generator (which I haven't
run in a while) off the list of candidate rules and my 419 corpus.
> yes and no, as JH works on them frequently.
Correct. And changes can affect the scoring.
> Hopefully John has some time & patience and we can agree on a basic set of
> rules and their scores.
The basic set of subrules changes, and I'm open to additions, but the mtea
for the ADVANCE_FEE rules is generated by a GA process.
I don't have a problem with statically scoring them, in fact that's what
the static sandbox score file was intended to address. I don't know if
that's the _best_ way...
I too would like to see a way to assign a minimum score. The GA rescorer
seems to do some very counterintuitive things at times, and it would be
conforting to have a way to control it a little better.
-- John Hardin KA7OHZ http://www.impsec.org/~jhardin/ jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79 ----------------------------------------------------------------------- You know things are bad when Pravda says we [the USA] have gone too far to the left. -- Joe Huffman ----------------------------------------------------------------------- 3 days until Bill of Rights day